Transition to Thompson Microphysics for Microwave All-sky Assimilation #743

azadeh-gh · 2024-05-07T14:46:48Z

Description
Transition from using the GFDL microphysics scheme, currently in operation, to the Thompson scheme (GFSv17) is proposed for the microwave all-sky assimilation. This transition aims to enhance the accuracy and reliability of forecasting. Issue#719

Resolves #719

Type of change

Please delete options that are not relevant.

Bug fix (non-breaking change which fixes an issue)
New feature (non-breaking change which adds functionality)
Breaking change (fix or feature that would cause existing functionality to not work as expected)
This change requires a documentation update

How Has This Been Tested?

Checklist

My code follows the style guidelines of this project
I have performed a self-review of my own code
I have commented my code, particularly in hard-to-understand areas
New and existing tests pass with my changes
Any dependent changes have been merged and published

Thompson_MP_modifications

ease enter the commit message for your changes. Lines starting

RussTreadon-NOAA

Check for GSI Code Standard compliance. Minor items noted. Did not fully check crtm_interface.f90.

src/gsi/crtm_interface.f90

src/gsi/general_read_gfsatm.f90

src/gsi/netcdfgfs_io.f90

RussTreadon-NOAA

Thank you, @azadeh-gh , for updating the code. It looks much better now in terms of GSI code standards.

RussTreadon-NOAA · 2024-05-08T20:26:16Z

src/gsi/crtm_interface.f90

+ + sb(7)*tc0*tc0*cse(1) + sb(8)*tc0*cse(1)*cse(1) &
+ + sb(9)*tc0*tc0*tc0 + sb(10)*cse(1)*cse(1)*cse(1)
+ smoc = a_ * smob**b_
+ reff(k) = MAX(2.51E-6_r_kind, MIN(0.5*(smoc/smob), 1999.E-6_r_kind))


0.5 should be 0.5_r_kind

RussTreadon-NOAA · 2024-05-13T16:54:33Z

Rebuilt azadeh-gh:feature/ThompsonMP_Azadeh at c24e568 on Cactus and rerun ctests. Channels 1 to 13 of gmi_gpm now also shows contrl versus updat differences. AMSU-A differences occur in channels 1 to 5 and 15. ATMS differences occur in channels 1 to 6 and 16 to 22.

emilyhcliu · 2024-05-19T23:28:36Z

src/gsi/crtm_interface.f90

@@ -1730,6 +1761,7 @@ subroutine call_crtm(obstype,obstime,data_s,nchanl,nreal,ich, &
 ges_qsat (ixp,iy ,k, itsigp)*w10+ &
 ges_qsat (ix ,iyp,k, itsigp)*w01+ &
 ges_qsat (ixp,iyp,k, itsigp)*w11)*dtsigp
+ rh(k) = q(k)/qsat(k) !added for cloud fraction computation


@azadeh-gh The variable qsat is declared but not defined.

I think qsat should qs

emilyhcliu · 2024-05-20T00:23:35Z

src/gsi/crtm_interface.f90

@@ -1730,6 +1761,7 @@ subroutine call_crtm(obstype,obstime,data_s,nchanl,nreal,ich, &
 ges_qsat (ixp,iy ,k, itsigp)*w10+ &
 ges_qsat (ix ,iyp,k, itsigp)*w01+ &
 ges_qsat (ixp,iyp,k, itsigp)*w11)*dtsigp
+ rh(k) = q(k)/qsat(k) !added for cloud fraction computation


I think qsat should qs

emilyhcliu · 2024-05-20T01:30:13Z

src/gsi/crtm_interface.f90

+ cf_calc = zero
+ ! sum of the cloud condensate amount for all 5 hydrometeos (ql + qi + qs + qg + qh + qr) 
+ qcond(:) = cloud(:,1) + cloud(:,2) + cloud(:,3) + cloud(:,4) + cloud(:,5)
+ call calc_thompson_cloudfrac (nsig , lmfdeep2, xrc3, prsl, qcond(:), rh, qsat,cf_calc)


Same here as above for qsat.

emilyhcliu · 2024-05-20T16:51:41Z

src/gsi/crtm_interface.f90

+ if (trim(cloud_names_fwd(ii))=='ni' .and. atmosphere(1)%temperature(k)<t0c) then
+ cloud_cont(k,ii)=max(1.001_r_kind*1.0E-6_r_kind, cloud_cont(k,ii))
+ cloud_efr(k,ii)=max(5.001_r_kind, cloud_efr(k,ii))
+ endif
+ if (trim(cloud_names_fwd(ii))=='nr' .and. atmosphere(1)%temperature(k)<t0c) then
+ cloud_cont(k,ii)=max(1.001_r_kind*1.0E-6_r_kind, cloud_cont(k,ii))
+ cloud_efr(k,ii)=max(5.001_r_kind, cloud_efr(k,ii))
+ endif
+


These lines are not needed and should be removed.
We do not need to check number concentration since CRTM does not take number concentration as input. We just need number concentration (ni and nr) for effective radius calculation

emilyhcliu · 2024-05-28T02:46:52Z

@azadeh-gh So far, we have made sure the code can handle imp_physics=8 and imp_imp_physics=11.
The last thing to check is the effective radius and cloud fraction from Thompson and GFDL.
I ran both cases and printed out all atmospheric profiles (T, q, ql, qi, qr, qs, qg, cloud fraction and effective radius of all hydrometeors) from both methods (Thompson and GFDL). I do not think the effective radius from the Thompson scheme is calculated correctly. Please check the subroutines (calc_thompson_reff and calc_thompson_cloudfrac) again.

They can be found on HERA:
/scratch2/NCEPDEV/stmp1/Emily.Liu/GSI382/thompson-debug-amsua
/scratch2/NCEPDEV/stmp1/Emily.Liu/GSI382/gfdl-debug-amsua

Look at the 10th profile in fort.1004102 for both cases (search icmask in the file until you see the first icmask is T (true))

 icmask        :  T
 lat lon       :   -57.3952000000000        106.477000000000
 channel no.   :            2
 emissivity    :   0.527780801232015        206.827145270325
tsim tsim_clr :    1.747967983657567572E+02   1.614273477412949944E+02   8.695621635317450915E-01
 tcc           :   0.869562163531745

It should be very easy to find the bug since we have GFDL as a reference. The values (effective radius) from Thompson should not be too different.

emilyhcliu · 2024-05-31T17:46:57Z

@azadeh-gh I performed a sanity check for two branches: develop (control) and feature/ThompsonMP_Azadeh (experiment)

GSI Builds, Scripts, and IC for the Sanity Check
The set up can be found on HERA:
control: /scratch1/NCEPDEV/da/Emily.Liu/git/GSI/GSI-emilyhcliu/GSI
experiment: /scratch1/NCEPDEV/da/Emily.Liu/git/GSI/GSI-emilyhcliu/GSI-thompson-azadeh

The scripts to run the two experiments:
/scratch1/NCEPDEV/da/Emily.Liu/git/GSI/GSI-emilyhcliu/scripts/rungsi_develop.sh
/scratch1/NCEPDEV/da/Emily.Liu/git/GSI/GSI-emilyhcliu/scripts/rungsi_thompson_azadeh_for_develop.sh

The initial conditions:
COMROOT=/scratch2/NCEPDEV/stmp3/Azadeh.Gholoubi/TestExperiment/MonthThompsonMP

Sanity Check Configuration
The sanity check has the following configurations for 2023050200 and identical results are expected.

same initial condition (with the Thompson forecast; so the cloud profile has number concentrations for rain (nr) and ice (ni))
same set of observing system
4denvar run
default global anavinfo for the control; update global anavinfo (with nr and ni added) for the experiment
imp_physics is set to 11 (GFDL) for both control and experiment
use GFDL cloud table for both control and experiment

Test Run and Output Directory
Control: /scratch2/NCEPDEV/stmp1/Emily.Liu/GSI382/develop
Experiment: /scratch2/NCEPDEV/stmp1/Emily.Liu/GSI382/thompson-azadeh-for-develop

Results
I checked the Jo tables and the minimization step output for all 202 steps. They are identical between the Control and Experiment (GSI develop vs. your branch). The evidence is attached below.
The regression test for your branch against the develop should be passed.

Control at the end of the first minimization loop/ second outer loop

 36:  Begin J table inner/outer loop           0           2
 36:     J contribution                 1             2             3             4             5             6             7
 36: background              1.008904E+05  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00
 36: excess   moisture       1.369165E-01  1.374345E-01  1.363932E-01  1.343722E-01  1.326088E-01  1.305097E-01  1.280860E-01
 36: surface pressure        1.914876E+03  1.204426E+03  1.277209E+03  2.650365E+03  1.194631E+03  1.212828E+03  1.259412E+02
 36: temperature             2.554828E+03  5.320637E+03  8.836285E+03  1.734312E+04  1.034742E+04  5.817320E+03  3.132768E+03
 36: wind                    8.087878E+03  2.197073E+04  2.761711E+04  4.641993E+04  3.415848E+04  2.043609E+04  9.243730E+03
 36: moisture                1.995685E+02  5.340923E+02  1.496204E+03  2.176303E+03  4.961715E+02  3.997539E+02  4.190078E+02
 36: sst                     3.793319E+02  3.853462E+02  3.804480E+02  4.858384E+02  4.747631E+02  4.413564E+02  5.844765E+01
 36: ozone                   1.391325E+03  2.662759E+03  2.652272E+03  1.791333E+03  2.379520E+03  2.215390E+03  1.913433E+03
 36: gps bending angle       1.715324E+04  3.820486E+04  3.878015E+04  3.878843E+04  3.869788E+04  4.171309E+04  2.063999E+04
 36: radiance                8.382490E+04  1.490322E+05  1.658015E+05  1.494925E+05  1.490969E+05  1.598365E+05  7.196505E+04
 36:      J term                                     J
 36: background                   1.0089041321553968E+05
 36: excess   moisture            9.3632088894809418E-01
 36: surface pressure             9.5802762609576021E+03
 36: temperature                  5.3352382572843781E+04
 36: wind                         1.6793395224131059E+05
 36: moisture                     5.7211010589933849E+03
 36: sst                          2.6055316431712367E+03
 36: ozone                        1.5006030751918539E+04
 36: gps bending angle            2.3397761775828814E+05
 36: radiance                     9.2904948129338946E+05
 36:  -----------------------------------------------------
 36:  J Global                    1.5181177231173015E+06

Experiment at the end of the first minimization loop / second outerloop

4452  36:  Begin J table inner/outer loop           0           2
4453  36:     J contribution                 1             2             3             4             5             6             7
4454  36: background              1.008904E+05  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00
4455  36: excess   moisture       1.369165E-01  1.374345E-01  1.363932E-01  1.343722E-01  1.326088E-01  1.305097E-01  1.280860E-01
4456  36: surface pressure        1.914876E+03  1.204426E+03  1.277209E+03  2.650365E+03  1.194631E+03  1.212828E+03  1.259412E+02
4457  36: temperature             2.554828E+03  5.320637E+03  8.836285E+03  1.734312E+04  1.034742E+04  5.817320E+03  3.132768E+03
4458  36: wind                    8.087878E+03  2.197073E+04  2.761711E+04  4.641993E+04  3.415848E+04  2.043609E+04  9.243730E+03
4459  36: moisture                1.995685E+02  5.340923E+02  1.496204E+03  2.176303E+03  4.961715E+02  3.997539E+02  4.190078E+02
4460  36: sst                     3.793319E+02  3.853462E+02  3.804480E+02  4.858384E+02  4.747631E+02  4.413564E+02  5.844765E+01
4461  36: ozone                   1.391325E+03  2.662759E+03  2.652272E+03  1.791333E+03  2.379520E+03  2.215390E+03  1.913433E+03
4462  36: gps bending angle       1.715324E+04  3.820486E+04  3.878015E+04  3.878843E+04  3.869788E+04  4.171309E+04  2.063999E+04
4463  36: radiance                8.382490E+04  1.490322E+05  1.658015E+05  1.494925E+05  1.490969E+05  1.598365E+05  7.196505E+04
4464  36:      J term                                     J
4465  36: background                   1.0089041321553968E+05
4466  36: excess   moisture            9.3632088894809418E-01
4467  36: surface pressure             9.5802762609576021E+03
4468  36: temperature                  5.3352382572843781E+04
4469  36: wind                         1.6793395224131059E+05
4470  36: moisture                     5.7211010589933849E+03
4471  36: sst                          2.6055316431712367E+03
4472  36: ozone                        1.5006030751918539E+04
4473  36: gps bending angle            2.3397761775828814E+05
4474  36: radiance                     9.2904948129338946E+05
4475  36:  -----------------------------------------------------
4476  36:  J Global                    1.5181177231173015E+06

Here are the final minimization step evaluation:
Control

cost,grad,step,b,step? =   2 150  1.465160993399310159E+06  1.971038208820891313E+01  2.822385197809975099E 01  1.604897083342762043E+00  good

Experiment

cost,grad,step,b,step? =   2 150  1.465160993399310159E+06  1.971038208820891313E+01  2.822385197809975099E-01  1.604897083342762043E+00  good

azadeh-gh and others added 11 commits August 15, 2023 21:12

Thompson_MP_modifications

7613041

Merge pull request #1 from azadeh-gh/ThompsonScheme_Azadeh

4b67e28

Thompson_MP_modifications

Merge branch 'NOAA-EMC:develop' into develop

8930f93

Merge branch 'NOAA-EMC:develop' into develop

5584613

Merge branch 'NOAA-EMC:develop' into develop

1244005

Merge branch 'NOAA-EMC:develop' into develop

1d98cdf

Merge branch 'NOAA-EMC:develop' into develop

2e4320e

Merge branch 'NOAA-EMC:develop' into develop

e9cdff0

Modified code based on FV3jedi code

7c6bc78

Corrected g

3a197ca

ease enter the commit message for your changes. Lines starting

Merge branch 'NOAA-EMC:develop' into feature/ThompsonMP_Azadeh

ae089db

azadeh-gh mentioned this pull request May 7, 2024

Transition to Thompson Microphysics Scheme for Microwave All-sky Assimilation #719

Open

RussTreadon-NOAA added this to the GFS v17 GSI/EnKF development milestone May 7, 2024

RussTreadon-NOAA assigned azadeh-gh May 7, 2024

RussTreadon-NOAA requested review from emilyhcliu and xincjin-NOAA May 7, 2024 14:58

RussTreadon-NOAA self-assigned this May 7, 2024

RussTreadon-NOAA self-requested a review May 7, 2024 15:00

RussTreadon-NOAA reviewed May 7, 2024

View reviewed changes

azadeh-gh and others added 2 commits May 7, 2024 22:54

Merge branch 'NOAA-EMC:develop' into feature/ThompsonMP_Azadeh

d37f8eb

Addressed comments, Followed GSI code Standards

7623bac

RussTreadon-NOAA reviewed May 8, 2024

View reviewed changes

Updated prt_guess.f90 to print ni and nr in stdout

c24e568

emilyhcliu suggested changes May 19, 2024

View reviewed changes

emilyhcliu suggested changes May 20, 2024

View reviewed changes

azadeh-gh and others added 2 commits May 21, 2024 03:34

Applied Emily's changes

f91f63f

Merge branch 'NOAA-EMC:develop' into feature/ThompsonMP_Azadeh

3e7126f

Updated the crtm_interface code for thompson reff subroutine

abed302

azadeh-gh and others added 2 commits June 6, 2024 15:23

adding lmfdee2 = .true. for cloud fraction calculation

59b19a8

Merge branch 'NOAA-EMC:develop' into feature/ThompsonMP_Azadeh

582be1d

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Transition to Thompson Microphysics for Microwave All-sky Assimilation #743

Transition to Thompson Microphysics for Microwave All-sky Assimilation #743

azadeh-gh commented May 7, 2024 •

edited by RussTreadon-NOAA

Loading

RussTreadon-NOAA left a comment

RussTreadon-NOAA left a comment

RussTreadon-NOAA May 8, 2024

RussTreadon-NOAA commented May 13, 2024

emilyhcliu May 19, 2024

emilyhcliu May 20, 2024

emilyhcliu May 20, 2024

emilyhcliu May 20, 2024

emilyhcliu May 20, 2024

emilyhcliu commented May 28, 2024

emilyhcliu commented May 31, 2024 •

edited

Loading

Transition to Thompson Microphysics for Microwave All-sky Assimilation #743

Are you sure you want to change the base?

Transition to Thompson Microphysics for Microwave All-sky Assimilation #743

Conversation

azadeh-gh commented May 7, 2024 • edited by RussTreadon-NOAA Loading

RussTreadon-NOAA left a comment

Choose a reason for hiding this comment

RussTreadon-NOAA left a comment

Choose a reason for hiding this comment

RussTreadon-NOAA May 8, 2024

Choose a reason for hiding this comment

RussTreadon-NOAA commented May 13, 2024

emilyhcliu May 19, 2024

Choose a reason for hiding this comment

emilyhcliu May 20, 2024

Choose a reason for hiding this comment

emilyhcliu May 20, 2024

Choose a reason for hiding this comment

emilyhcliu May 20, 2024

Choose a reason for hiding this comment

emilyhcliu May 20, 2024

Choose a reason for hiding this comment

emilyhcliu commented May 28, 2024

emilyhcliu commented May 31, 2024 • edited Loading

azadeh-gh commented May 7, 2024 •

edited by RussTreadon-NOAA

Loading

emilyhcliu commented May 31, 2024 •

edited

Loading